The trial is based on preliminary data showing that upright tilt table testing can provoke a drop in blood pressure consistent with neurally mediated hypotension (NMH) in a high proportion of those with chronic fatigue syndrome (CFS), and that unblinded treatment of the NMH leads to an improvement in CFS symptoms in 40-70% of CFS patients. The specific aim of this study is to determine whether patients aged 18 to 50 years with CFS and NMH will have a greater improvement in (1) self- reported general sense of well being and (2) objective orthostatic tolerance when treated with fludrocortisone than when treated with placebo. Eligible subjects are randomized to receive either fludrocortisone (escalating to 0. I mg/day) or placebo for nine weeks. In week 8-9 of treatment, subjects undergo repeat tilt testing. The primary outcome measure is the proportion with a 15 point improvement in wellness on a 100 point wellness score, and a secondary outcome is the proportion with improvement in the number of minutes before the development of hypotension during upright tilt. The first patient was studied as part of this protocol 4 March 1996. From that date until 30 November 1997, we evaluated 124 patients with chronic fatigue syndrome. In 72 of these (58%), hypotension was documented during the first two stages of upright tilt testing. One patient with hypotension during the tilt test had a wellness score that was above 70 at the time of the positive tilt test, and was therefore ineligible for participation. Two other patients with positive tilt tests were not randomized due to concerns about their mental health. A fourth patient with a positive tilt test was not randomized after urinalysis at NIH showed that the patient had not been honest about the medications she was taking and was therefore ineligible for the study. The remaining 68 subjects were randomized to fludrocortisone or placebo. Of these, 7 subjects have dropped out due to side effects. Two subjects did not complete all paperwork as required and did not return for follow-up tilt testing due to unrelated illness, and were therefore counted as drop outs. Additionally, 1 subject withdrew from the study due to fear of venipuncture and I subject withdrew due to the amount of paperwork required. The total number of subjects counted as drop outs is 11. The study procedures appear to be well tolerated, and there have been no unanticipated problems with tilt testing or with the study medications. With very few exceptions, the study outcome data have been received in a timely manner by the company coordinating data entry. Of those randomized, 47 (69 %) are female. 70 subjects enrolled are Caucasian, 2 subjects are black. No Hispanic, Asian or Pacific Islander, or American Indian or Alaskan Native has applied to the study. These demographic characteristics reflect the epidemiologic patterns seen in other studies of CFS.